Formal Recursion
A finite seed and non-terminating recursive rule can generate a history space with no fixed finite upper bound.
PDF summary and reader page
A conditional Surtea-Austin correspondence note that separates generated information, recovered information, physical support, carrier stress-energy, and the stronger conjecture of an independent information stress-energy sector.
This page has been built from the local PDF source. It belongs directly after the RSG and ITR correspondence note because it takes the next step: it asks what has to be true before information can matter physically, gravitationally, or observationally.
The document is careful about claim status. It does not say that abstract information gravitates. It says that information becomes physically relevant through a support, a recovery channel, and an energy-bearing carrier, unless a separate action-level information sector is explicitly supplied.
Best first use. Use this page as the claim-discipline bridge between information-theoretic language and stress-energy language. It keeps "recoverable structure" separate from "extra substance."
The paper has two levels. Level A is conservative: recoverable information has gravitational relevance through the stress-energy of the carrier on which it is physically encoded. Level B is conjectural: a recovered-information density may define its own stress-energy tensor only if it is placed in an action with independently fixed support, coupling, units, coefficients, conservation rules, and failure conditions.
abstract information != T_munu
generated information
-> support
-> boundary and closure coupling
-> recovery
-> physical measure
-> stress-energy
That chain is the heart of the document. It says where the bridge is allowed to operate and where it stops if the required physical or operational pieces are missing.
The introduction divides the paper into three distinct claims. Keeping them separate prevents a common overclaim: treating unbounded formal generation as if it were already physical storage, measurement, or stress-energy.
A finite seed and non-terminating recursive rule can generate a history space with no fixed finite upper bound.
A finite observer with a finite recovery vector recovers only finite information in finite observation time.
Information becomes gravitationally relevant through an energy-bearing carrier or through a separately defined information action.
If support, coupling, observable density, or conservation cannot be specified, the strong bridge fails back to the carrier-only claim.
Generated information is a formal measure over the set of histories available after repeated recursion. Recovered information is what a finite channel can actually extract. The paper treats these as different objects, not as two names for the same quantity.
I_gen(n) = log2 |H_n|
if |H_n| >= b^n, then I_gen(n) >= n*log2(b)
I_rec(t_obs | V_rec) <= C_R * t_obs
The first line is formal growth. The last line is an operational recovery bound. A bounded observer can fail to recover information without proving that the generated history space has ended, and a large generated history space does not imply that a finite system stores an explicit infinity.
The support layer supplies the topological grammar. A universe is represented as U = (M,D),
where D is a partition of M. The terms "interior", "closure", and "boundary"
are computed relative to that partition. In this paper, a boundary is not just a drawn edge. It is a
statement about which partition cells are partially involved and therefore available for interaction
or recovery.
int_D(X) = union of partition cells fully contained in X
cl_D(X) = union of partition cells that intersect X
bd_D(X) = cl_D(X) - int_D(X)
This matters because recovered information must be encoded on some support. The support is not the whole physical state, but it is the place where encoded structure can meet boundary, closure, coupling, measurement, and physical measure.
sigma_n = (X_n, phi_n, mu_n, S_n)
support != whole physical state
The recovery vector is the document's operational gate. It records the medium, resolution, sampling, coupling, tolerance, ordering time, and decoder required for a system to turn encoded structure into recovered information.
V_rec = (m_med, r_res, nu_samp, kappa, delta, tau_ord, D_dec)
I_rec(O | V_rec)
= D_dec o kappa o Pi_delta o S(r_res, nu_samp, tau_ord) o m_med(O)
The chain is type-safe: the measurement medium produces a signal space, the sampling and tolerance create a recoverable pattern, coupling makes the pattern available, and the decoder returns recovered information. If the recovery vector is orthogonal to the object, recovered information is approximately zero without implying that the encoded structure itself is absent.
V_rec perpendicular O -> I_rec(O | V_rec) ~= 0
The paper then connects the recovery discussion back to the RSG represented measure. Generated histories are filtered by accumulated effective loss, producing survival weights and then normalised representation weights.
A_i(t) = integral_0^t Lambda_i(tau) dtau
S_i(t) = exp(-A_i(t))
Z(t) = sum_j exp(-A_j(t)), 0 < Z(t) < infinity
p_i(t) = S_i(t) / Z(t) = exp(-A_i(t)) / Z(t)
p_i / p_j = exp[-A_i + A_j]
H_surv = -sum_i p_i log(p_i)
N_eff = exp(H_surv)
These are representation weights. They say which histories remain live under the survival filter. They are not automatically ontic probabilities unless an additional interpretation is supplied.
The conservative route is simple and strong. A recoverable message does not gravitate because of its abstract meaning. It is physically relevant because it is encoded in a carrier, and the carrier has stress-energy by the usual action principle.
T_munu = -2/sqrt(-g) * delta S_car / delta g^munu
I_rec != 0 -> kappa != 0 -> physical carrier coupling
if mu_n contains S_car[g,psi], then T_car_munu carries the gravitational relevance
This keeps the bridge grounded. The information has to be encoded in something. The gravitational part belongs to the energy and momentum of that something unless a stronger information-sector action is separately defined.
The strong conjecture is not dismissed, but it is made expensive. A recovered-information density
iota_R(x) can only define an information stress-energy tensor after an action has been
supplied. That action needs units, coefficients, transformation behaviour, conservation or exchange
laws, and operational measurement rules.
iota_R(x) = 0 if x notin cl_D(X)
iota_R(x) = 0 if kappa(O,V_rec) = 0
epsilon_I = chi_I * iota_R
Landauer specialization: chi_I = k_B * T_L * ln(2)
S_I[g,iota_R,X,kappa,V_rec]
= integral sqrt(-g) * L_I(g,iota_R,X,kappa,V_rec) d4x
T_I_munu = -2/sqrt(-g) * delta S_I / delta g^munu
The note gives templates rather than predictions: potential-like, gradient-field, radiation-like, and matter-like forms are possible only after the missing definitions are fixed before comparison with observations.
If an information sector is added to the Einstein equation, conservation cannot be optional. The Bianchi identity forces either total conservation or an explicit exchange current between the carrier and information sectors.
G_munu = (8*pi*G/c^4) * (T_car_munu + T_I_munu)
nabla_mu T_car^munu = -Q^nu
nabla_mu T_I^munu = Q^nu
The bridge fails if it lacks support, boundary and closure relation, recovery channel, observable density, action, conservation or exchange law, equation of state, or coefficients fixed before the target observation. This is one of the most useful parts of the paper: it gives the idea a testable perimeter.
Light is the clean example because the carrier is already familiar. A modulated electromagnetic field can carry a recoverable message, but its stress-energy is the Maxwell field's stress-energy. The message changes the readable structure of the field, not the rule by which the field gravitates.
E = h*f
p = E/c
E_total = sum_i N_i*h*f_i
P_total = sum_i N_i*(h*f_i/c)*n_hat_i
V_IE = c^2 * P_total / E_total
For ideal coherent vacuum light, V_IE = c*n_hat. In a medium, the effective transport
vector is closer to the group velocity. Either way, transport velocity is a carrier property; recovered
information rate is a channel property.
The paper makes a sharp distinction between propagation speed and recoverable bit-rate. The light cone bounds signal transport, but it does not assign a universal number of bits per second to every light beam. Bit-rate depends on bandwidth, modulation, power, photon flux, detector resolution, coupling, decoder, and noise.
dI_rec/dt <= C_R
tau_bit = Delta_x / c
B = 1/tau_bit = c / Delta_x
light speed sets transport velocity; channel capacity sets bit-rate
The 532 nm green-laser example is used to show the trap. One bit per optical cycle and one bit per
photon are possible scale comparisons, but neither is automatically the real communication capacity.
A chosen modulation example at 1.0e6 bits/s makes each bit occupy about 300 m
of the beam and contain about 2.68e9 photons.
Material channels change propagation velocity, attenuation, dispersion, and coupling. They still do not create a universal informational speed limit. The recoverable rate remains a property of the whole recovery vector.
v_g = c / n_g
v_g = 1 / sqrt(L_prime * C_prime)
P_rx = P_tx * exp(-alpha_mat * L)
C_R,mat = Delta_f * log2(1 + SNR)
B_real <= C_R,mat
ell_bit = v_g / B_real
The vinyl example treats a groove as an analogue recovery channel whose effective rate depends on stylus contact, cartridge coupling, bandwidth, equalisation, decoder, and noise. The copper example treats signalling as guided electromagnetic propagation: conductor geometry and dielectric set group velocity, while bandwidth and noise set capacity.
The worked optical example uses a 1 mW, 532 nm beam over a square-millimetre scale area with a 1 Mbit/s modulation. The point is not that the gravitational contribution is large. It is that the same beam that carries a recoverable message also carries energy, momentum, pressure, and a tiny mass-equivalent.
E_gamma = h*c/lambda
N_dot = P_beam / E_gamma
I_beam = P_beam / A_beam
u = I_beam / c
p_rad = I_beam / c
I_rec(t_obs | V_rec,laser) <= B_mod * t_obs
The recovered information depends on modulation, detector bandwidth, coupling, decoder, and noise. The stress-energy belongs to the electromagnetic carrier. The bridge is therefore real but modest: recoverable optical information is structure in an energy-bearing field.
The gamma-light analogy is conditional. Gamma photons do not become matter merely because they are light. Matter production requires sufficient energy, the correct interaction channel, conservation of energy-momentum, and boundary or recoil support.
gamma + gamma -> e_minus + e_plus
p_gamma1^mu + p_gamma2^mu = p_minus^mu + p_plus^mu
s >= 4*m_e^2*c^2
physical encoding + recovery coupling + energy-momentum support -> stress-energy contribution
The analogy says that transitions require closure. In the same spirit, information becomes stress-energy-relevant only when encoding, support, coupling, and energy-momentum support are all present.
The paper ends by separating the observable limit from an existence limit. The observable universe is limited by horizon, finite propagation, and finite recovery channel. That does not by itself define the whole universe or a global update rate.
observable universe != whole universe
observable limit = horizon + finite propagation + finite recovery channel
finite recovery != finite generated history
Local waves can have frequencies, but the full field history is a spectrum, propagation history, and boundary-interaction record. Hard numerical ceilings should therefore be treated as recovery limits unless an operational existence limit is supplied.
These compact formulas carry the page's usable spine. They are written in plain text so they can be copied into notes without image extraction.
I_rec(t_obs | V_rec) <= C_R * t_obs
sigma_n = (X_n, phi_n, mu_n, S_n)
p_i(t) = exp(-A_i(t)) / sum_j exp(-A_j(t))
H_surv = -sum_i p_i log(p_i)
T_munu = -2/sqrt(-g) * delta S / delta g^munu
S_I = integral sqrt(-g) * L_I d4x
speed sets transport; capacity sets recoverable bit-rate
The safe reading is: this document supplies a conditional correspondence framework. It is strong because it says exactly what must be supplied before an information-stress-energy claim is meaningful. It is cautious because, without those supplied pieces, the theory remains at the carrier route.
c alone; specify the recovery channel.This is reading-order item 15. It follows the RSG and ITR correspondence note because it sharpens the bridge from information language into physical measure language. It comes before the black-hole recovery note because it supplies the support, boundary, recovery, and stress-energy discipline used there.
Read sections 2 through 8 of the PDF first: claim status, generated versus recovered information, support, recovery vector, survival measure, carrier route, and independent sector. Then read the worked light, vinyl, copper, laser, and gamma sections as examples of the same rule: information is recoverable structure on a supported carrier, not a free-floating addition to energy.